Okay, I apologize ahead of time... I am a GM not an Engineer

[The Beave]

Sandy,

I have worked with a former engineer, who now works at Orban, He is the most down to earth guy and knows the whole system (as well as others) and most likely could help you with one phone call.

Call or e-mail Kevin Clayborn @ Orban/CRL USA. I think he can do alot of wonders for you. Sometimes good information comes from the manufacturers. (without wanting to upgrade you unless you need it.)

Good Luck,
The Beave

 

[LA_Guy]

What kind of processing is used in front of the 160kbps stream? Sounds really good!

I still, however, would always prefer never to have my over the air revenue dependent upon a stream via the internet.

Scares me. Really.

Just a db or so of very gentle AGC. The player is Foobar 2000 v .9.5.3 & I'm using their Vlevel plug in very gently.

This stream has been up for 8 weeks from Hollywood, CA to Vermont and hasn't missed a beat.

I'm even using a 'defective' Instreamer that has no analog audio capability whatsoever (blown audio chip).

I'm feeding the Instreamer via coaxial SP/DIF.

To illustrate my point, the sources are a mixture of FLAC files, and MP3s of 160, 192 and 265 kbps. You tell me which is which!

 

[menotti1]

well my golden ears CAN'T tell which is which.DAM GOOD audio.Sounds alot better than most 950 analog STL's,i;ve been around.Like it or not AUDIO OVER IP is the new platform for broadcast and just audio period. Barix just happens to be the most cost effective.

 

[busyradioguy]

LA-Guy,

With all due respect, I think you're a little too focused on my comments regarding audio quality, at the expense of my comments about the delay issues. The delay alone, is the primary reason I would not use IP Streaming as an STL.

Some of you guys who are claiming IP Audio as an STL is becoming the norm, I must ask. What were you using as an STL before the days of IP Streaming?

Also, it was mentioned that the delay in HD radio was reason enough to stop monitoring the off air signal when on live. So long as the analog channel remains, monitoring the off air signal while you are on live, is really the only way to go. The only exception might be for a talk show using a profanity delay.

 

[bossbill]

I find it a sad state of affairs that "broadcast engineers" are suggesting that lossy data compression has a purpose in quality broadcasting, especially since so many radio programs are satellite-delivered in the first place, which of course uses compression.

Audio delay is not the issue. That can be handled with separate audio processors in studios so the jock hears something that is acceptable in the headphones; the issue is that we have become way too tolerant of degraded audio. Just because some of us can't hear the difference doesn't mean that listeners aren't subliminally tuning out due to listener fatigue.

By the time the record producers get done totally ruining what could have been a clean recording, and we download it as an mp3, then load it into automation, run it through processing and clipping, and send it to the transmitter through an IP scheme of any sort, no wonder we're losing listeners.

 

[menotti1]

what is sad but good is most listeners don;t notice or can't tell the difference.i'm waiting for the barix 1000.will use it for STL .PCM uncompressed with ip radio.But in reality the flac or mp3 256 would work fine.My prediction is the Barix 1000 will sell like mink coats on dollar days,just as i predicted for the Omnia One FM.My preference will be PCM.I have always suggested non linear when feasible.Radio may be losing listeners,but content is the main cause not delivery.(just plain Vanilla radio,same 300 songs over and over,dull , no personality,etc..)

 

[LA_Guy]

I find it a sad state of affairs that "broadcast engineers" are suggesting that lossy data compression has a purpose in quality broadcasting, especially since so many radio programs are satellite-delivered in the first place, which of course uses compression.

Audio delay is not the issue. That can be handled with separate audio processors in studios so the jock hears something that is acceptable in the headphones; the issue is that we have become way too tolerant of degraded audio. Just because some of us can't hear the difference doesn't mean that listeners aren't subliminally tuning out due to listener fatigue.

By the time the record producers get done totally ruining what could have been a clean recording, and we download it as an mp3, then load it into automation, run it through processing and clipping, and send it to the transmitter through an IP scheme of any sort, no wonder we're losing listeners.

Whether any one of us likes it or not, lossy coding is here to stay. Watch any TV? MPEG2 coding. Listen to an IPOD or other player? MP3 or WMA. If you follow the rules, this stuff can sound pretty damn good. I think you should listen before you comment.

 

[Sgeirk]

I have listened. I was downloading mp3's off the internet in 1998. Noticed immediately that they weren't ready for prime time.

You can GET AWAY with using compression on audio on the radio...but if you could choose otherwise, why would you?

Just because TV is doing it and people are downloading mp3's doesn't mean that I want to sound like a TV or an mp3. I agree with Bill W wholeheartedly. I'm still more concerned with sounding better than the other 50 radio stations in the market.

I go back to my earlier comment...just because you can do something, doesn't always mean you should.

 

[LA_Guy]

LA-Guy,

With all due respect, I think you're a little too focused on my comments regarding audio quality, at the expense of my comments about the delay issues. The delay alone, is the primary reason I would not use IP Streaming as an STL.

Some of you guys who are claiming IP Audio as an STL is becoming the norm, I must ask. What were you using as an STL before the days of IP Streaming?

Also, it was mentioned that the delay in HD radio was reason enough to stop monitoring the off air signal when on live. So long as the analog channel remains, monitoring the off air signal while you are on live, is really the only way to go. The only exception might be for a talk show using a profanity delay.

I was using a 950 composite STL with a s/n of 62 db, that if the wind was blowing in the right direction passed a noise proof.

If you are running IBOC, you MUST delay the analog the same 8 seconds as the digital is! Otherwise, when the radios switch back and forth between them (and car radios do, all the time) the listener isn't totally thrown off.

 

[Chuck]

Ahhh did you know that one of the oldest and most respected (by most) engineers in Dallas is Sandy Sanford???? (Long time member of SBE Chapter 67)........thought you might want to know

Well, I know the Sandy Sanford in Dallas, and I don't think this is the same person. That doesn't make his questions any less valid though....

 

[kyscott]

I find it a sad state of affairs that "broadcast engineers" are suggesting that lossy data compression has a purpose in quality broadcasting, especially since so many radio programs are satellite-delivered in the first place, which of course uses compression.

Audio delay is not the issue. That can be handled with separate audio processors in studios so the jock hears something that is acceptable in the headphones; the issue is that we have become way too tolerant of degraded audio. Just because some of us can't hear the difference doesn't mean that listeners aren't subliminally tuning out due to listener fatigue.

By the time the record producers get done totally ruining what could have been a clean recording, and we download it as an mp3, then load it into automation, run it through processing and clipping, and send it to the transmitter through an IP scheme of any sort, no wonder we're losing listeners.

All your points are true. What are we to do about it? I used the Barix boxes for an STL for one of my FMs. When I flip it back and forth between my program loop, you can't tell the difference. It sounds good, it works and it's cheap. As reliable as the Barix units are proving themselves to be, the $800 program loop from Bellsouth will be going away by the end of the year in favor of the $40 DSL line.

 

[busyradioguy]

L.A._Guy,

OK you got me on that delay introduction on the analog side, for HD. I had forgotten all about that. ;D

However, I am one of those who has lost faith in HD Radio, simply because it isn't taking off. That's likely because there's no video with the audio and also due to the fact that HD radios are being given the back-burner treatment by a lot of retailers. I've heard too many horror stories from folks who have actually gone shopping to evaluate HD radio.

 

[FredRichards]

Unlike the TV "HD" side of things, where digital television replicated the service area, I'm finding a big difference between the analog coverage and digital coverage. I'm afraid some folks have been sold a large lemon.

 

[TomT]

Hey, Mr. G.M., see where this has gone? :

Get three engineers in a room, and you have six opinions....

If you can scrape up the $2400, buy an Omnia One and replace the 8100. Somebody in your neighborhood has a tunable modulation monitor they can bring over to set your levels. The folks at Omnia are very helpful in helping you twiddle with the settings to get the sound you want. Try their preset settings first, you may just want to leave well enough alone. I went from an 8100 to an Omnia 3t (predecessor to the Omnia One) on a classic rock--my GM has been very happy since (except..."I know it is at 100%, but couldn't you make it just a little louder...?")

The Barix system they are arguing over is around a grand--but you need to have a broadband connection, so you are going to pay rent to someone. Composite systems from TFT or Armstrong run around $4500 list--and you own the path. RF licensing will do your paperwork for $400. If you have a long path, call someone like V-Soft for a path analysis. Again, a little more up-front money, but once you are done--you're done!

 

[kyscott]

The Barix system they are arguing over is around a grand--but you need to have a broadband connection, so you are going to pay rent to someone. Composite systems from TFT or Armstrong run around $4500 list--and you own the path. RF licensing will do your paperwork for $400. If you have a long path, call someone like V-Soft for a path analysis. Again, a little more up-front money, but once you are done--you're done!

I bought an Instreamer AND Exstreamer for $350. I pay $40 a month for my DSL line at the transmitter, using the existing high speed internet service at the studio. It will take a little over 100 months for the RF link to pay for itself using those numbers. Also, that's assuming you can find a frequency. AND assuming there is nothing wrong with your path. AND assuming you can get a path. AND assuming zoning laws will allow you to put up a tower high enough to get a path. For whatever reason, if I need to move my studios, I don't have to relicense or restudy a path. Audio over IP is getting too cheap to ignore.

 

[menotti1]

Just waiting for the Barix 1000 to hit the market.i prefer PCM. a pair of IP radios and i will be good to go.yep.audio over IP is the new frontier.

 

[TomT]

kyscott--he has a path...it's now discrete. He wants to add RDS, presumably song and title, which is most easiest done at the studio. If the path is short enough, he probably can just drop in a composite link.

As soon as Marti gets me a functioning receiver, I will be rebuilding a composite link that's been in place for about 25 years. Still using the original 505C system. Don't think there are that many digital links that have been around for as long unchanged.

 

[OKCRadioGuy]

A lot of people speak favorably of the Barix Instreamer....and while I've never installed one, I pushed for a compsoite STL system and was over-ruled due to money issues. The Barix "won"...the station "lost". The Barix has butchered the sound of the station to the point that I am ashamed of it and regularly distance myself from it. You can no longer listen to yourself on the air live, nor can two physically separate studios interract live on a broadcast (due to the 1 to 20 second variable delay) and every time there's an ISP issue, you're off the air. For those who are using those things with success, I am genuinely happy for you. But my experience says if you have a working STL, the Barix option is a huge downgrade.

I'd much rather use a real STL than anything like a Barix. With that being said, I have a staiton going in likely in the next couple of years that's about 60 miles away from me. Our statewide internet provider for the state has a point of presence in both my studio and near the tower site. I will be using Barix equipment (hopefully the 1000s that CAN be ran linear if the bandwidth will allow it, or mp3 if it doesn't) to get it down to the other school where they net connection will be, then STLing it out from there to the tower. Since it's all within the same fiber-feed system with microwave reduancies (can you say overbuilt ) it should be pretty damn reliable, hopefully moreso than Ma Bell. The cost should be better too. Barix units certainly have their place, but I most whole hardely agree good old 950MHz STLs are the best really.

Let me put a little plug in here for J-Squared. Sandy would do well to give him a buzz and see if he'll send her some loners where she could send the dual system back in for updating and re-capping. If she kept that system in place (after referb) and got a nice Omnia One for the processor I bet she'd be astounded how good the station would sound, and all for a very reasonable price.

 

[busyradioguy]

The Barix system they are arguing over is around a grand--but you need to have a broadband connection, so you are going to pay rent to someone. Composite systems from TFT or Armstrong run around $4500 list--and you own the path. RF licensing will do your paperwork for $400. If you have a long path, call someone like V-Soft for a path analysis. Again, a little more up-front money, but once you are done--you're done!

I bought an Instreamer AND Exstreamer for $350. I pay $40 a month for my DSL line at the transmitter, using the existing high speed internet service at the studio. It will take a little over 100 months for the RF link to pay for itself using those numbers. Also, that's assuming you can find a frequency. AND assuming there is nothing wrong with your path. AND assuming you can get a path. AND assuming zoning laws will allow you to put up a tower high enough to get a path. For whatever reason, if I need to move my studios, I don't have to relicense or restudy a path. Audio over IP is getting too cheap to ignore.

Don't you see though? You're ripping yourself off. You're having to pay for the Internet service. What is your backup plan when the IP Service dies? What happens to your budget if your IP Service changes hands, and increases your bill?

You're never going to recoup your loses by paying $40 / month for that DSL, against the investment of a microwave system. $40 a month times 12 months a year is $480. $480 times say 5 years is $2,400. That is assuming your IP provider doesn't increase your monthly fee. And they very well may, if you are using their network as a 24 / 7 STL system. That's a hell of a lot of bandwidth on their network.

 

[gtomas]

When you talk about build up using but one system, you show your ignorance of lossy coding. YES you can have build up-but ONLY if you use DIFFERENT coding methods and/or rates. If you start out with a 44.1 kHz 256 kbps MP3 file and maintain THAT coding and bit rate (or better), then you will NOT have any further quality loss-all things being equal. What WILL make a difference are many D/A and A/D conversions-especially with poor quality codecs (like the ones in most consumer computers)

I'm sorry but this is completely WRONG! MP3 is a lossy codec. As such it psychoacoustically reduces audio information to fit the target bitrate. Note the word reduces - the process is irreversible. Destructive if you want. Once removed, audio information can't be retrieved again.

When such audio is subjected to another coding the same process is repeated. It doesn't matter what sample rate or bitrate the second codec runs at. It still has to another reduction all over again. It will remove even more audio information (from the audio that already has large part of audio information removed).

If you do it again, the process repeats. This is very well known as codec cascading and has been studied. Now, it is true that it has been observed that if you keep the same bitrate and coding, there is somewhat less degradation. But quite far from claim that there is no quality loss. With each lossy coding process, there is and always will be quality loss.

Getting back to the discussion, I agree that the final coding is something we can't escape any more. However, we can and I should avoid any coding prior to final transmission coding to keep the audio quality the best possible. We should also fight for the highest possible bitrate for that final transmission coding, no matter how efficient coding is used.

That being said, I would recommend an uncompressed digital STL as a best solution for getting your audio from the studio to the transmitter. If that is not possible, analog STL is second preferred choice. My third choice would be a low compression time domain algorithm such as apt-x. Only if even that is not possible I would look into coding such as AAC (let me remind you AAC is 30% more efficient than MP3 and has slight better properties when cascaded). My personal last choice would be an MP3 running at anything below 320/256kbps...


Regards,
Goran Tomas